Font Size: a A A

The Application Of Multivariate Analysis Methods For The Determination Of Mixed Dye Concentration

Posted on:2013-11-10Degree:MasterType:Thesis
Country:ChinaCandidate:J LiFull Text:PDF
GTID:2231330362975160Subject:Biochemical Engineering
Abstract/Summary:PDF Full Text Request
At present, the dyeing and finishing enterprises can’t determine to the concentrationof the dye liquor in dyeing process interiorly. In this field it is still in the initial stage inChina, The mixture of dye liquor concentration determination of aspect needs to be furtherexploration and research, expecting to be able to quickly and accurately analysis dyeliquor concentration in order to improve the work efficiency and ensure the quality ofproducts of dyeing and finishing enterprises.In this article, adopting several spectrophotometric methods on two components ofmixed dyeing concentration determination. At the same time partial least square method isapplied to direct dyes and reactive dyes concentration determination. Building the forecastmodel of used of individual component concentration prediction.Forecasting the actualconcentration change of dyeing liquor in dyeing process. The results show that partial leastsquare method combined with photometric method is applied to dye liquor concentrationdetermination is feasible.Research mainly includes the following four aspects:(1)For two components of mixed dyeing, adopting the maximum wavelength-simultaneousequations method, dual-wavelength spectrophotometry and three-wavelength spectrophotometry todetermine various component dyes concentration. The results show that the three methodscan be used to determine two components mixed dyeing concentration, and they haverapid, accurate of characteristics.(2)For three components of mixed dyeing, first make sure the direct dyes and reactivedyes concentrations of linear range. Three direct dyes concentration linear range are0.01g/L-0.1g/L.The each components of linear correlation coefficients is more than0.99.Three reactive dyes concentration linear range are0.01g/L-0.06g/L. The eachcomponents of linear correlation coefficients is more than0.99too. In linear range ofconcentration, compounding the different concentration of the three components mixeddyeing samples respectively.30samples for correction set,10samples for forecast set.Using UV-Vis spectrophotometer to acquire calibrating samples absorption spectrumcurve and forecast set sample absorption spectrum curve.(3)Establishing the mixed dyeing of concentration quantitative prediction models forthe direct dyes and the reactive dyes by the partial least squares method. Eachcomponents of the direct dyes models of linear correlation coefficients is0.996、0.998、0.988; RMSEC respectively is0.0013、0.0003、0.0041; RMSEP respectively is0.018、0.0007、0.0129. Various components model of reactive dyes of the linear correlationcoefficient is0.9985、0.9970、0.9659; RMSEC respectively is0.00083、0.0012、0.0035,prediction error within10%.(4)Built model is used for dyeing process of dye liquor concentration prediction andmap the dyeing rate curve and dye concentration change with time curve in the process of dyeing. Using standard curve gain concentration to compared with model forecastconcentration in order to inspect the forecast accuracy of model. The result shows that themodel can well predict dyes concentration change in dyeing process.Dyeing rate accordswith the actual dyeing. It provides a effective way to partial least square method forapplied to the on-line monitoring...
Keywords/Search Tags:Partial Least Squares, Mixed dyeing concentration, Spectrophotometry, Dyeing process
PDF Full Text Request
Related items